1. Field of the Invention
This invention relates to a method for selectively separating live cells which have expressed a specific gene.
2. Related Background Art
In the case that the translation products of a gene are cell surface molecules, a method for selectively separating cells which have expressed the specific gene, while being viable, is to make fluorescence labeled antibodies bind to the surface molecules for labeling the cells fluorescently, to identify fluorescing cells by flow cytometry, and to separate the identified cells with a cell sorter (Fluorescence Activated Cell Sorter, FACS). In addition, the panning method is also known wherein only the objective cells are absorbed on the bottom surface of a dish over which is covered with antibodies specifically binding to the cell surface molecules.
In the case that the translation products of a gene are not cell surface molecules but localize in the cells (in the cytoplasm or in organella), the method described above cannot be adopted. In this case, it is theoretically possible to fluorescently label the gene-expressing cells by introduction of fluorescence-labeled antibodies that are specific to the molecules localized in the cells into the cells, through microinjection and to separate the objective gene-expressing cells with the cell sorter described above based on the difference in fluorescence intensity of the cells with irradiation of laser beam or the like.
However, for the cell labeling method by microinjection described above, the method can not label many cells at once. The number of the cells to which the labeled antibodies can be introduced for one experiment is at most ten or less. In addition, it is not easy to introduce the solution of a polymer with the molecular weight greater than 120,000 like an antibody with high concentration into the cell because of its high viscosity. Therefore, microinjection is impractical to label the sufficient number of objective cells efficiently.
In the case that the translation products of the gene are not cell surface molecules, but molecules that are liberated into the extracellular fluid and that do not remain in the cell or near the cell membrane, it is very difficult to selectively trap the molecules to separate the cells expressing the specific gene, from other molecules with the approaches described above. This is because, during the process where polypeptide chains generated based on the genetic information are folded and secreted, their structure changes gradually and from time to time to prevent any known antibodies from binding to the polypeptide chains within or on the surfaces of live cells efficiently. Also, even in the case that the translation products are present on the surfaces of cells, it is difficult to selectively separate the cells unless the molecules are specifically present on the surfaces of particular cells.
A typical example, where the situation described above exists and when it is difficult to separate the objective live cells selectively, includes the case where cells secreting a specific cytokine are selectively separated using the cytokine as a selection marker.
When an antigen invades an organism, helper T cells (CD4+ T cells) that recognize the antigen as a foreign matter are activated, and then they will be differentiated into TH1 and TH2, which have different immune functions from each other: TH1 (T Helper 1) which is responsible for cellular immune functions, e.g., activation of macrophages to remove foreign matters by phagocytosis; and TH2 (T Helper 2) which has humoral immune functions, e.g., activation of B cells to produce antibody molecules to neutralize foreign matters and (See FIG. 94). TH1 and TH2 produce cytokines, interleukin-2 (IL-2) and interleukin-4 (IL-4), respectively. In the healthy state, TH1 and TH2 control each other's functions and keep a balance. However, once this relationship is disrupted, it causes various infections or autoimmune disorders.
If TH1 or TH2 can be selectively separated and obtained, it will be medically important, because their application can be contemplated in supplementing immune functions or the like. Thus, a variety of attempts have been made to find molecules that are present on the surface of TH1 or TH2, that can be used for their separation and obtaining.
For example, it has been reported that the tissue infiltration which is dependent on adhesive molecules, P- and E-selectin, is observed specifically with human TH1 (Austrup, F. et al. Nature, 385, 81–83, 1997). This suggests that ligands adhering specifically to the selectins are present on TH1 cell surfaces. However, when reactivity for P- and E-selectin is examined by flow cytometry, the results are TH1:TH2=131:52 for P-selectin, and TH1:TH2=668:88 for E-selectin; therefore, the specificity is not complete. These results can be interpreted as reflecting the fact that particularly notable P- and E-selectin ligand expression is induced in TH1 under physiological conditions unique to inflamed tissues (such as skin and joints).
Receptors for CC-chemokine (CCR3), eotaxin, have also been reported to be present with approximate specificity on human TH2 cell surfaces (Sallusto, F. et al. Science, 277, 1997). However, since CCR3-negative T cell groups also include IL-4 producing TH2 cells in a proportion of 1.9%, the specificity is not complete. Furthermore, the presence of many more of the same receptors on eosinophil and basophil cell surfaces than on TH2 raises the risk of possible contamination by cells other than TH2 if CCR3+ cells are simply separated from T lymphocytes that have been crudely purified from blood.
The receptor CCR5 for other CC-chemokines such as MIP-1β and IP10 and the receptor CXCR3 for the CXC-chemokine SDF-1 have been reported to be present with approximate specificity on human TH1 cell surfaces (Loestscher, P. et al., Nature, 391, 344, 1998). However, since one of the nine TH2 clones obtained here was CCR5+, the specificity is not complete. Furthermore, while TH1 shows higher CXCR3 gene expression and CXC-chemokine dependent migration than TH2, CXCR3 gene expression was also confirmed in all of the TH2 clones examined, and therefore, the specificity is not complete. Moreover, CCR5 and CXCR3 are also present on neutrophil cell surfaces, and therefore the risk exists of possible contamination by neutrophils in CCR5+ or CXCR3+ cells separated from T lymphocytes that have been crudely purified from blood.
In addition, IL-12 (interleukin-12) receptor (IL-12R) has been reported to be present with approximate specificity on human TH1 cell surfaces (Rogge, L. et al., J. Exp. Med., 185, 825, 1997). However, while TH1 cell surfaces bear high affinity receptors (Kd value=27 pM) and low affinity receptors (Kd value=5 nM) for IL-12 at 140 molecules and 450 molecules per cell surface, respectively, similar low affinity receptors (Kd value=2 nM) are also present on TH2 cells at 200 molecules per cell surface. This means that IL-12R cannot be used as a definitive TH1 cell surface marker. Moreover, since IL-12R is also present on the cell surfaces of NK cells, the risk exists of possible contamination by NK cells in IL-12R positive cells separated from T lymphocytes that have been crudely purified from blood.
IL-18 (interleukin-18) receptor (IL-18R) is another receptor reported to be present specifically on the cell surface of a TH1 clone established from transgenic mice with T cell receptors for ovalbumin (Xu, D. et al., J. Exp. Med., 188, 1485, 1998). However, like IL-18R, the ST2L molecule belonging to the interleukin-1 receptor (IL-1R) family is also known to be present on TH2 cell surfaces. Because gene homology within the IL-1R family is particularly high in humans, IL-18R cannot be considered a definitive cell surface marker in humans and no reports have yet been published on their presence specifically on TH1 cell surfaces. Also, since IL-18R is much more abundantly present on monocyte, neutrophil and NK cell surfaces than on TH1, the risk exists of possible contamination by cells other than TH1 in IL-18R positive cells separated from T lymphocytes that have been crudely purified from blood.
The reports cited above suggested that receptors for cytokines, chemokines and the like present on TH1 or TH2 cell surfaces vary considerably in terms of amount (number of per cell surface) and quality (affinity of the receptors for their ligands or intracellular transduction of stimuli upon binding to ligands), and that the distribution of such receptors therefore highly favors either TH1 or TH2. The reason for the favorableness of cell surface molecules toward either TH1 or TH2 is believed to arise from the biological environment (physiological conditions) surrounding the helper T cells.
For example IL-12, which is one of the ligands for these cytokine receptors, is a cytokine secreted by macrophages, etc. upon initial infection (Walker, W. et al., J. Immunol., 162, 5894, 1999), and IL-18 is also known to be produced by activated macrophages and Kupffer cells (Yoshimoto, T. et al., J. Immunol., 161, 3400, 1998). Naturally, both of these cytokines have more connections with TH1 than with TH2, in light of the cellular immunity function of the former, and they are considered to be factors that perform transduction of physiological information from macrophages to TH1 (i.e., that activate TH1 in the body).
This suggests a connection between activation of macrophages by TH1 and reception of stimuli (IL-12 and IL-18) returned from macrophages, whereby macrophages activated by TH1 eliminate foreign matters in the body while also activating TH1. As this interdependent relationship is established at sites of inflammation in the body, it is fully expected for the number of receptors for IL-12 and IL-18 and the activity of the receptor molecules to increase significantly on TH1 cell surfaces. Further, since TH2 cells are not exposed to the same conditions in the body as TH1, it is surmised that they have no need to receive IL-12 or IL-18. However, as long as IL-12 or IL-18 receptors are detected even slightly on TH2 cell surfaces, it cannot be denied that TH2 also has the potential to respond to IL-12 or IL-18.
It is therefore inconceivable that these cytokine receptors are definitive markers that can distinguish TH1 from among TH1 and TH2. In addition, since these chemokine and cytokine receptors that are predominantly distributed on TH1 and TH2 cells are also found distributed among other cell types such as NK cells, they are considered impractical as markers for distinguishing TH1 or TH2 from each other in blood samples. For example, cell specimens containing CD4+ cells (helper T cells) that are separated and purified by common methods from blood samples taken from humans usually include contamination by monocytes and granulocytes, and these false positive cells may be expected to be mistaken for TH1 or TH2.
As stated above, it is, therefore, very difficult to selectively separate TH1 and TH2 cells based on surface molecules. Moreover, since the cytokines (IL-2 and IL-4) produced by TH1 and TH2 do not remain in the cell or near the cell membrane but are liberated into the extracellular fluid, it is difficult to selectively separate TH1 and TH2 using these cytokines.